DESCRIPTION (APPLICATION ABSTRACT): Several recent reports have highlighted growing concerns about the quality of patient care in the U.S. (IOM, 2001; Kizer, 2000; Kohn, Corrigan & Donaldson, 2000; Schuster et al., 1998). While these concerns are not new, there is a perception that the quality of patient care is decreasing. At the same time, injuries to health care workers remain a serious problem. As it is likely that reductions in staffing and other organizational changes may act as risk factors for adverse outcomes for both patients and workers, patient and worker outcomes will be linked by institution to address this crucial hypothesis. To do this, patient and worker injury rates in both acute and long term care facilities will be assessed with an ecological framework, measured over time, using the following specific aims: 1) Examine the relationship of adverse patient outcomes (post-surgical complications, post-surgical misadventures) to staffing variables (staffing, skill mix, RN to total) at the institutional level; 2) Examine the relationship of adverse worker outcomes (worker injuries) to staffing variables at the institutional level; 3) Identify changes in the rate of adverse patient outcomes in relation to staffing variables by institution, over time (1998-00); 4) Identify changes in the rate of adverse worker outcomes in relation to staffing variables by institution, over time (1998- 00); and 5) Examine the organizational characteristics of hospitals and nursing homes in relation to rates of adverse patient and worker outcomes. Three years of administrative data (1998-2000) will be compiled from three states using acute care hospitals and nursing homes. Acute care adverse patient outcomes will come from the HCUP QI and for nursing homes from the HCFA Minimum Data Set. Organizational descriptors will be obtained from the AHA Annual Survey database (e.g. total FTEs, RN FTE, nonprofit status, and size) and OSCAR (beds, special services, and staffing). Worker injuries will be obtained from state First Report of Injury databases. Longitudinal trends over three years will be examined, using (Generalized Estimating Equations, in addition to cross-sectional analyses. This study will provide empirical data to support the utility of modifications in the organization of health systems, which are designed to improve the quality and safety of patient care while maximizing the health and productivity of workers.